Dynamic Discounting Redefined: AI’s Role in Global Vendor Relationships

A. N. Srinivasan, Vice President at SRF Limited, took the audience on an intriguing journey beyond the realms of traditional AI applications during the Machine Learning Developers Summit (MLDS) 2024. Delving into the financial domain, Srinivasan shed light on leveraging Oracle’s data mining capabilities to enhance working capital management. In this unconventional exploration, he unveiled how AI could optimize discounts, credit periods, and vendor relationships.

Exploring New Frontiers: AI in Finance

In a departure from conventional AI narratives, Srinivasan emphasized the power of AI in finance. Drawing from his extensive 28-year career, notably 21+ years specializing in Oracle applications, he highlighted the critical role of working capital in organizational success. Srinivasan’s groundbreaking contributions, acknowledged with the Zenith AI Award, set the stage for a captivating discourse on AI’s potential in financial strategies.

Dynamic Discounting: Transforming Vendor Relationships

Collaborating with global partner C24, a dynamic discounting expert, SRF embarked on a mission to optimize working capital. The dynamic discounting model allowed vendors worldwide to choose between discounts, credits, or financing through the bank. Srinivasan explained how AI was the missing piece in transforming supplier relationships. The goal was to forecast which vendors would offer discounts, providing a competitive edge in procurement.

AI-Powered Data Mining: The Oracle Advantage

Srinivasan shared insights into their strategic choice of Oracle data mining. Leveraging the in-built algorithms, he navigated through the data exploration, anomaly detection, and model-building phases. Oracle’s data mining capabilities streamlined tasks such as exploratory data analysis and data cleaning, paving the way for effective model training.

Deciphering Model Accuracy: The Oracle Data Mining Journey

Highlighting the significance of data attributes, Srinivasan showcased the rigorous attribute selection process. The audience gained insights into the sampling process and the importance of attributes like product patterns, vendor invoices, and payment histories. The selection of algorithms—Random Forest, Support Vector Machine, and Naïve Bayes—was meticulously explained, emphasizing the need for accuracy and precision in financial predictions.

The Critical Step in AI Success

The presentation underlined the criticality of deploying AI models successfully into production. Srinivasan stressed the importance of selecting the right platform for deployment, ensuring a seamless and continuous flow of data for model training. Whether using Oracle data mining or other platforms, the success of AI in a production environment is pivotal to its real-world impact.

Transformative Impact on Vendor Forecasting

Srinivasan unveiled the transformative impact of the AI model on vendor forecasting. The model identified prospective vendors likely to offer discounts, streamlining procurement efforts. What started as an internal initiative for SRF soon extended its benefits to global partner C24. Within three months, C24 reported profits, underscoring the tangible and collaborative success of the AI model.

Choosing the Right Algorithm

The decision-making process behind choosing Random Forest as the preferred algorithm was demystified. Srinivasan shared that real-world confirmation from C24 validated Random Forest’s predictions, solidifying its selection. The alignment of the model’s forecast with practical business outcomes showcased the intricate interplay between AI technology and real-world dynamics.

Conclusion

In conclusion, Srinivasan emphasized that the true success of AI lies in its ability to translate predictions into tangible business impacts. While AI models may offer numerous insights and attributes, their real value is realized when they contribute to an organization’s bottom line. This unique foray into AI’s financial applications showcased the collaborative success of SRF and C24, proving that AI’s transformative potential extends far beyond traditional boundaries.